Tephrofacts and the First Human Occupation of the French Massif Central Jean-Paul Raynal, Lionel Magoga, Peter Bindon

Tephrofacts and the first human occupation of the French Massif central Jean-Paul Raynal, Lionel Magoga, Peter Bindon

To cite this version:

Jean-Paul Raynal, Lionel Magoga, Peter Bindon. Tephrofacts and the first human occupation of the French Massif central. The Earliest Occupation of Europe. Proceedings of the European Sci- ence Foundation, Workshop at tautavel, France, 1993, edited by Wil ROEBROEKS and Thus Van KOLFSCHOTEN, 1995, Leiden, Netherlands. pp.129-146. halshs-00004086

HAL Id: halshs-00004086 https://halshs.archives-ouvertes.fr/halshs-00004086 Submitted on 21 Jul 2005

HAL is a multi-disciplinary open access L’archive ouverte pluridisciplinaire HAL, est archive for the deposit and dissemination of sci- destinée au dépôt et à la diffusion de documents entific research documents, whether they are pub- scientifiques de niveau recherche, publiés ou non, lished or not. The documents may come from émanant des établissements d’enseignement et de teaching and research institutions in France or recherche français ou étrangers, des laboratoires abroad, or from public or private research centers. publics ou privés. European Science Fondation Meeting , Tautavel , sous presse - 1 -

TEPHROFACTS AND THE FIRST HUMAN OCCUPATION OF THE FRENCH MASSIF CENTRAL

Jean-Paul RAYNAL(1), Lionel MAGOGA(2), Peter BINDON(3)

Mots-clés : Massif Central, tephrofacts, pétrographie, Paléolithique ancien.

Key-words: France, Massif Central, Tephrofacts, Petrography, Lower Palaeolithic.

Résumé: Les auteurs examinent la production par le volcanisme de pseudo-artéfacts. L'existence de ces téphrofacts doit être prise en compte pour l'analyse des outillages réputés très anciens. Dans l'état actuel des recherches, seul le site de Soleilhac offre les meilleures garanties d'une présence humaine très ancienne (0,5-0,6 Ma ?) dans le Massif Central français.

Abstract: This paper examines the production of pseudo-artefacts by vulcanism and discusses their presence in supposedly ancient sites in the French Massif-Central. At this time, only the site of Soleilhac offers sufficient evidence for very ancient human occupation of the region (0.5-O.6 Ma ?).

(1) Université de Bordeaux 1, Institut du Quaternaire, UMR 9933 CNRS, Avenue des Facultés, F-33405 Talence Cedex. (2) Résidence Raymond, 48 Allée des Ailes, F-03200 Vichy. (3) Department of Anthropology, Western Australian Museum, Francis Street, Perth 6000, Western Australia. European Science Fondation Meeting , Tautavel , sous presse - 2 -

Introduction

The production of geofacts, objects resulting from natural fracture and imitating artefacts (HAYNES, 1973), is a phenomenon long recognized in many different sedimentary contexts (PEI, 1937; MORTELMANS, 1947; BREUIL, 1955; CLARK, 1958; BOURDIER, 1967; FOURNIER, 1971; RAYNAL et TEXIER, 1989; RAYNAL et al, 1990). A number of works analyse these, particularly for flint, and try to elaborate a method which distinguishes clearly between intentionally flaked objects and products of nature (BOULE, 1889, 1905, 1921; HAWARD, 1911; MOIR, 1911; WARREN, 1914; GRAYSON, 1986; SCHNURRENBERGER and ALAN, 1985; PEACOCK, 1991).

The Massif Central in the central mountainous region of France experienced active vulcanism during the Miocene (15-20 M.a) and several glacial events during the Pleistocene. Because vulcanism and frost action generate rock fracturing, there is thus a high probability of the natural occurrence of pseudo-artefacts in this area. Discovery of flint pseudo-tools confused a number of archaeologists during the nineteenth century beginning with TARDY (1869), who reported the discovery of Miocene eoliths from Le Puy Courny in the Cantal. More than fifty years afterwards, MARTY was still arguing for this interpretation (CAPITAN et MARTY, 1924).

The Massif Central region is well known for important excavations from the Upper Palaeolithic and has considerable archaeological potential for sites yielding material from the Lower and Middle Palaeolithic. During the last twenty years, much research has been undertaken in the region towards discovering traces of the first human occupation of Europe. These investigations have been concentrated in Velay, a province to the south of Auvergne which is rich in Plio-Pleistocene fauna occurring in between volcanic sediments which offer the possibility of obtaining a long sequence of palaeomagnetic and radiometric dates. Basse-Auvergne and Bourbonnais have yielded a number of classic Acheulian bifaces lacking stratigraphic context and in surface complexes of the Allier tools made on quartz cobbles have been discovered.

The lithic series discovered in the Massif Central, principally in Velay, has been classified under the name "Most Ancient Palaeolithic" (BONIFAY and BONIFAY, 1983) and has been proposed as resolving the question of the time of the initial human occupation of Europe (BONIFAY, CONSIGNY, LIABEUF, 1989; BONIFAY, 1981, 1983, 1989a, b, c, 1991). This perspective, which is of considerable conceptual importance and thus must be founded on decisive arguments, is not accepted unanimously by the scientific community, as doubts have been expressed in many publications (DELSON, 1989; BOEDA, 1990; VILLA, 1991; FARIZY in DIAZ, 1993).

We report here some results of an examination of the production of geofacts by vulcanism. We will refer to these as tephrofacts. These pseudo-artefacts are fashioned in materials other than flint which occur in the local environment of supposedly ancient archaeological sites. Our observations, begun in 1989, are founded on the European Science Fondation Meeting , Tautavel , sous presse - 3 -

examination of a number of volcanic structures and the deposits and epiclastites associated with these structures (figure 1). These investigations were aimed at elucidating the possible presence of hominids in the Upper Pliocene and Lower Pleistocene of this part of France. Special attention has been given to the site of Blassac and part of the abundant series of supposed artefacts discovered in an ancient context there and also to the site of Soleilhac, yielding the strongest evidence for an ancient occupation of Velay.

1 - THE PRODUCTION OF TEPHROFACTS

The production of pseudo-artefacts or tephrofacts in volcanic contexts, is known to have occurred in pyroclastites of the Eifel in Germany (BOSINSKI et al., 1986). We collected some tephrofacts on the site of Kärlich and in strombolian lapilli of the Schweinskopf volcano in the company of G. BOSINSKI in 1992. These tephrofacts superficially resemble human artefacts and their differentiation from the latter is difficult when they are discovered apart from the primary volcanic context. (KULEMEYER, 1986).

In Velay, many Pleistocene maar tuff-rings (La Sauvetat, Les Farges, Saint-Eble, Soleilhac, Senèze, Blassac/Les Blanches), basanitic breccias (Sainte-Anne) and one strombolian cone have yielded tephrofacts with various petrological origins. Moreover, many other geofacts have been collected in a large number of volcanic sediments (Saint- Vidal, Vals, Brioude).

Among these are a number of flakes and some objects with multiple flake scars with apparent ordered intent: None of them would be discarded out of hand if they occurred in archaeological contexts. However, they are undoubtedly tephrofacts which have resulted from several mechanical and thermal actions during various different eruptive stages of volcanic events.

2- PETROGRAPHY OF THE TEPHROFACTS

The raw material of the tephrofacts collected come from the regional basement or from the Plio-Pleistocene sedimentary formations which have been altered by the volcanic eruptions. Some of these materials still outcrop in the immediate environment of the volcanic formations investigated. The following diverse rocks have been identified : vein quartz, pegmatitic quartz, fine grained granite, oriented granite, migmatitic gneiss with sillimanite, lamprophyre and various basalts.

The Blassac-Les Blanches tuff-ring has yielded seventy-two tephrofacts. The distribution of petrographic types does not vary significantly between the chunks, flakes or pseudo-artefacts (figure 2). The strong representation of materials of mediocre quality for flaking (granites and gneiss) and the absence of flint is a distinct characteristic of this series. European Science Fondation Meeting , Tautavel , sous presse - 4 -

3 - NATURAL FLAKES

In this paper we do not discuss the classic "pot-lid" flake form well known to occur as a result of thermally induced reduction sequences. We note however the absence in the literature of observations concerning the characteristics of these thermal fracture features when they occur adjacent to the edges of irregular chunks. The natural flakes derived from these situations exhibit a pseudo-striking platform produced when the fracture surface intersects an adjoining edge. Although these objects may be superficially identified as humanly produced flakes, this position can be readily discarded for flint when none of the other features of humanly produced flakes can be identified (point of percussion, ventral bulbar scar, radial stress marks, etc).

Fifty natural flakes were collected for study from the tuff-rings of Blassac-Les Blanches, La Sauvetat and Soleilhac (figure 3). Some of them were in juxtaposition with the pseudo-core.

Of these, fifteen (30%) exhibit cortical striking platforms. These can be subdivided into those with a total or partial cortical dorsal surface extending to the edges(11) and those with non cortical edges (4). In the case of a series produced by intentional flaking, these objects would be considered of primary and secondary generation.

Flakes without cortical striking platforms represent 70% of the total (35 objects) and this group is composed of those with cortical dorsal surfaces (8) and those with non cortical edges (27). In a series produced by intentional flaking, these would be considered third generation flakes.

Pseudo-retouch sometimes appearing contiguous can be seen on numerous flakes and is comparable with the utilized or retouched edges of humanly produced artefacts (figure 3).

The proportion of types of naturally produced flakes in this series is very different from that obtained from our experiments with quartz pebbles. They also differ from those of the flaked quartz assemblages collected from the upper and middle terraces of the Allier in Bourbonnais: an assemblage of tephrofacts is characterized by an overrepresentation of flakes with non cortical striking platforms and non cortical edges and an underrepresentation of flakes with cortical striking platforms and edges totally or partially cortical (figure 4).

Pseudo-flakes have a weight below 150g in more than 96 % of cases and rarely exceed 500g (figure 5). A few exceptions exist, represented by very large individuals which weigh over 2000g.

The weight criterion is not a discriminating factor between tephrofacts and artefacts. The distribution of weight ratios of pseudo-flakes and flakes manufactured and recovered in primary position (whether in eruptive breccias, archaeological layers, or by experimentation) is identical. This distribution differs from those observed on collections from surface sites: for the flakes collected in Bourbonnais on the surface of European Science Fondation Meeting , Tautavel , sous presse - 5 -

the upper and middle terraces of the Allier for example, the weight ratio distribution can be explained by the removal of the smaller flakes through natural processes and selective collection (figure 5).

This natural organization of a succession of flake scars produces flakes of several types and results from repetitive uni-directional stresses, both mechanical and thermal.

4 - NATURAL OBJECTS WITH MULTIPLE FLAKE SCARS.

Events leading to the natural flaking of fragile source rocks are diverse in nature and may result in an apparently organized series of flake scars. The order in which these flakes are removed largely determines the morphology of each tephrofact. The organization of the natural flake scars on the blocks parallels the morphology of objects recognized among prehistoric archaeological assemblages.

The number of flake scars on the tephrofacts (between 1 and 22) is comparable with that observed on prehistoric objects and their random organization is comparable to the sequences observed on manufactured objects. However, contrary to what is evident on prehistoric objects, the detachment points of flakes observed on tephrofacts are very often impossible to determine and most flake scars occurr in random arrangements.

On 34 faces of objects with multiple flake scars (9 unifaces and 26 bifaces) there is a linear system of "working" arranged from right to left for three flake scars in two cases and in the remaining cases the working system is non linear with a maximum number of seven flake scars originating from one edge. In general, pseudo-working identified on the tephrofacts is mainly non-linear in arrangement.

The associations of flake scars on edges are shown in the following table:

Number of flake scars 1 2 3 4 5 6 7 Sample Unifacially "flaked" objects 1 2 1 3 0 2 0 9 faces Bifacially "flaked" objects* 2 3 6 5 4 3 4 27 faces *(each face is considered separatly)

Some natural objects appear similar to unifacially flaked choppers and bifacially flaked chopping tools in that they exhibit the transformation of a natural edge with at least two adjacent flakes (figure 6, n° 1, 2, 4, 5; figure 7, n° 1; figure 8, n° 5; figure 10, n° 2).

Natural multidirectional flaking around the periphery of various chunks and pieces produces objects which are similar to invasively flaked tabular cores and discoïdal cores (figure 7, n°s 3; figure 10, n° 3; figure 9, n° 1, 2). Sometimes the natural morphology is European Science Fondation Meeting , Tautavel , sous presse - 6 -

similar to that seen on bifacially flaked artefacts (figure 6, n° 3; figure 8, n° 1; figure 9, n° 3; figure 10, n° 1).

Other tephrofacts look similar to flaked polyhedrons. These globular objects have a large number of flake scars from which were removed flakes with a high number of non-cortical striking platforms (figures 7, n° 2, 4; figure 8, n° 3, 4; figure 10, n° 4; figure 11).

As if the distinction between tephrofacts and intentionally manufactured objects was not difficult enough, the case is further complicated by the poor raw materials in which they occur. This observation has already been discussed in other works attempting to distinguish between geofacts and artefacts (WATSON, 1968), even though some progress has been made in the case of flint (PEACOCK, 1991), a raw material for which there is today a considerable amount of information on its flaking properties.

5 - DISTINGUISHING BETWEEN TEPHROFACTS AND HUMANLY PRODUCED OBJECTS.

We have had recourse to a statistical approach to solve this problem, the details of which will be presented elsewhere. Between 24 and 62 tephrofacts, according to the number in each sample, have been added to a series of humanly produced lithics. The criteria used for describing the two groups were published by J. COLLINA-GIRARD (1975, 1986). We have discarded some of those criteria, calibrated some of his non-metrical observations and added some of our own (weight and number of flake scars).

Factorial analysis of correspondences reveals that the inclusion of the tephrofacts destroys the coherence of the assemblage. The presence of two discrete and distinct groups is readily observed in each series.

In the case of a principal component analysis, taking all the characteristics except the number of flake scars into account, no true distinction can be made between tephrofacts and humanly produced objects. The natural trend in tephrofacts is towards a spherical or sub-spherical shape, which most effectively resists further natural attrition. Human intervention interrupts this trend towards a globular form. However, industries containing a preponderence of polyhedrons cannot be discarded out of hand as having a natural origin and should be studied according to the scheme suggested above.

Parametric characteristics often used for studying a humanly produced lithic assemblage, when applied to tephrofacts, do not permit any distinction between the two. The results, rather than showing the differences, confirm the similarities between tephrofacts and manufactured objects, rendering the method useless. European Science Fondation Meeting , Tautavel , sous presse - 7 -

6 - REMARKS ON TWO REPUTEDLY ANCIENT SITES.

6.1 - BLASSAC-LES BATTANTS.

A series collected at Les Battants by F. CARRE (1978, 1983, 1991) is dated by a number of Potassium/Argon determinations on the basaltic lava-flow overlying the assemblage, which indicates an age around or beyond 2 M.a.. However, the fauna of Blassac-La Girondie, in an analogous stratigraphic position, seems to be much more recent, belonging to the Peyrolles biozone and probably dating from around 1.2-1.4 Ma (J. COUTHURES, 1982; F. CARRE et J. COUTHURES, 1982; J. COUTHURES et J.F. PASTRE, 1983; M. FOURIS, 1989; BONIFAY, 1991; CARRE, 1991).

A series of 278 lithic objects has been examined. This study is incomplete, however, because the matrix has been left on several objects. Observations to date are as follows :

- Most objects are on crystalline rocks with natural cleavage planes and the fractures in general follow these planes. This gives the pieces an appearance of being partly or intentionally flaked debitage. However, most of the flakes do not have normal feathered terminations.

- From the evidence of the flake scars, the angle of detachment approaches 90°, which is not usually encountered in intentional flaking.

- Several objects have gross crystalline irregularities which render them inappropriate raw materials for flaking.

- Flakes do not exhibit clear points of percussion.

- Some objects exhibit obvious thermal fractures.

- On many objects, the flake scars appear to originate from well outside the remaining volume. Objects of similar form, and the flakes removed from them by thermal action, have been collected from the tuff-ring of Blassac-Les Blanches. In several cases, these pseudo-nuclei and their conjoining flakes were found in primary position within the tuff with the flakes barely removed from the parent chunk.

- Several objects of very poor flaking quality show a number of repetitive flake scars, which is an unlikely occurrence even in a primitive series of intentionally flaked objects. Moreover, within the same stratigraphic level, materials with far better flaking properties are present.

- Finally, the petrographic nature of the series presents a distribution comparable to that of tephrofacts (figure 12).

The geological characteristics of the site of Blassac-Les Battants deserve a full scale discussion too extensive for this paper. We simply note here the undeniable European Science Fondation Meeting , Tautavel , sous presse - 8 -

inclusion of naturally fractured objects and the resemblance to the petrographic suite occupied by tephrofacts.

6.2 - CHILHAC III

The fossil locality of Chilhac III is known for a rich Villafranchian fauna (BOEUF, 1983) and lithic objects (GUTH, 1974; GUTH et BOEUF, 1977; GUTH et CHAVAILLON, 1985) which occur in a level whose age is presumed to be in the vicinity of 1.8-1.9 M.a., a date arrived at through palaeontological comparisons and extrapolation from absolute dates on a lava flow close to the site.

This stratigraphy exhibits deformations interpreted as resulting from solifluxion (TEXIER, 1985) or soil slips (CHAVAILLON, 1991). Many visits to the site during the excavations have convinced us of the existence of syn-sedimentary slumpings and removals. This clearly demonstrates dynamism of a kind known to occur on the margins of lacustrine systems and we postulate the likelihood of the existence here of an ancient phreatomagmatic structure. In this dynamic situation, the presence of relatively recent tools on quartz cobbles in adjacent localities (LE GALL et RAYNAL, 1986) renders likely a mixture of such objects with the ancient fauna.

We have not examined closely the complete series of objects discovered in a stratigraphic context (ensembles B to K) apart from those published by J. CHAVAILLON (1991). Except for the resemblance between retouched flakes and some tephrofacts (figure 3), we note that the petrographic nature of the objects offers a distribution identical to that of tephrofacts and of the series of pseudo-artefacts recovered at Blassac-Les Battants (figure 12).

A detailed examination of the site and its environment would clarify without doubt the archaeological and geological processes which resulted in the production of these objects and their association with the faunal remains.

7 - COMPARISON WITH THREE ARCHAEOLOGICAL SERIES.

We have chosen three archaeological series to illustrate the fundamental differences which exist between the petrography of tephrofacts and that of prehistoric tools from Velay. In the latter, the petrography clearly demonstrates human selection.

- At Nolhac, apart from exceptional flint objects, quartz was the principal material chosen (RIO CARRA, 1991; BONIFAY, 1991),

- In Soleilhac-Centre, quartz, basalt and flint are in that order the three dominant materials (BRACCO, 1991), European Science Fondation Meeting , Tautavel , sous presse - 9 -

- In level J1 in the cave of Sainte-Anne 1 at Polignac, provisionally reported at isotopic stage 6, basalts, phonolite, flint and quartz were in that order the materials worked.

The petrography of tools from the prehistoric sites is clearly different to that of the tephrofacts and is characterized (outside the constant presence of flint) by a choice of materials which does not include those with poor flaking characteristics like granites and gneiss (figure 13).

8 - CONCLUSIONS

8.1 - TEPHROFACTS ARE WIDELY DISTRIBUTED.

It is more than reasonable to assume that tephrofacts were produced in great numbers in the Massif Central during the numerous volcanic episodes which occurred since the Miocene. Without doubt, they were subsequently eroded out of their primary position and widely dispersed in the environment on numerous occasions.

The discovery in geological layers of some broken or apparently flaked pieces, flakes or objects exhibiting a more complex pattern of flake scars is therefore not considered a sufficient criterion for characterizing human activity, particularly in this region.

8.2 - NECESSARY RE-EXAMINATION OF SITES.

Sites considered indicative of human activity demonstrated by the presence of a series of supposedly humanly worked pebbles and cobbles demand a close scrutiny of all the available evidence, especially in the case of Blassac and Chilhac III.

Sites which have yielded a limited number of doubtful artefacts, tephrofacts or geofacts need further detailed discussion. This is the case for Perrier-Etouaires in Puy-de- Dôme (G.U.E.R.P.A., 1984), Saint-Eble (BONIFAY, 1989a), le Coupet (BONIFAY, 1989a) and La Roche-Lambert (BONIFAY, 1981) in Haute-Loire.

For some sites, association of lithic objects with fragmented faunal remains is not decisive anyway. The various models of fragmentation and preservation of the bones do not give at present sufficient evidence to identify any human involvement.

From the evidence, it must be recognized that there is considerable doubt concerning a very ancient human presence in the Massif Central.

In addition to the taphonomic approach, it is necessary to consider the following points carefully and systematically for all localities:

- Is there a possibility of pseudo-artefacts being produced? European Science Fondation Meeting , Tautavel , sous presse - 10 -

- If so, what are the likely characteristics of these objects ?

- Is there a geological explanation for the introduction of naturally flaked objects into the site ?

- Are lithic objects preferentially distributed in the supposed archaeological layer ? Is there a natural explanation for this ?

- Do conjoinable objects exist in the site, and do they occur adjacent to each other ?

- Are the flake scars observable on the objects arranged in a technologically "logical" sequence ?

- Are natural objects mixed with artefacts ?

- Are the faunal remains and the associated objects both in primary position ?

- Does the taphonomic history of the site provide any explanation for possible association of numbers of objects of different ages, for example reflecting periglacial phenomena ?

8.3 - OPPORTUNISTIC EXPLOITATION ?

Can we discard out of hand sites where only a few objects have been found ? The utilization of materials of poor flaking quality in these sites may simply be an example of human opportunism according to the law of least effort. Additionally we have the complicating factor among ancient assemblages of the likelihood of opportunistic exploitation of naturally fractured pieces which will occur anyway as a part of the site's assemblage. While these questions have already been discussed for several African archaeological sites, it is no doubt useful to consider the "advantage of thinking small as archaeology explores the most ancient spans of prehistory" (G. ISAAC et al. , 1981).

8.4 - THE TIME OF FIRST HUMAN OCCUPATION.

Only the sites where unquestionable archaeological elements have been discovered in a well documented long sequence should be considered to elucidate this problem. In the absence of precise dating elements for the site of Nolhac, only Soleilhac "the most recent of the very Lower Palaeolithic sites" (BONIFAY, 1991) remains to give an idea of the temporal remoteness of the human presence in Velay.

The age of Soleilhac has been established by biostratigraphic criteria (BONIFAY an BONIFAY, 1981), palaeomagnetic determinations (THOUVENY, 1983; THOUVENY and BONIFAY, 1984) and morphostratigraphy (BONIFAY and MERGOIL, 1988). Soleilhac- European Science Fondation Meeting , Tautavel , sous presse - 11 -

Centre has also been studied from a sedimentological and palynological point of view (RAYNAL, 1987 and unpublished). However, the fauna of Soleilhac belongs to the "transition fauna" which covers a broad time spectrum (BONIFAY, 1987). Furthermore, the sedimentological and palynological characters do not allow reliable dates to be determined and the paleomagnetism of the complete sequence has not been studied because of the unsuitability of some layers for the application of this method ( THOUVENY, op. cit., p. 81). Most interest has been generated by the tephrostratigraphy reported by A. TEULADE (1985, 1988, 1989) and determined from cores (BONIFAY and MERGOIL op. cit.).

A direct pumice ash-fall has been identified in the lake series underlying archaeological layer C. These tephra correlate with the "upper pumices of Sancy" (CANTAGREL and BAUBRON, 1983) and most particularly with the pumice of Neschers. A number of dating methods place the Neschers pumice around 0.8 M.a. (TEULADE, 1989, p. 145). More recently, however, a 40Ar/39Ar age of 0,58 ± 0.02 M.a. has been obtained (LO BELLO et al. , 1987), which has been confirmed by a thermoluminescence date of 0,52 ± 0,04 M.a. obtained by the quartz red peak TL method (PILLEYRE, 1991). Thus, the tephrostratigraphy contradicts the paleomagnetic data and necessitates consideration of a date around 0.5-0.6 M.a. for Soleilhac-Centre. This last date is in agreement with a late Cromerian age for the fauna, as suggested by VAN KOLFSHOTEN (ROEBROEKS and VAN KOLFSHOTEN, this volume).

8.5 - DIRECTIONS OF FUTURE RESEARCH.

If we discard all the doubtful ancient sites occurring in the Massif Central, the earliest settlement history of the region must be rewritten. However, all is not lost yet. The evidence for the earliest occupation of Auvergne and Velay can be revisited, evaluated and carefully dissected avoiding the traps of their geological context. Paradoxically, in the final analysis it would be this last which furnishes a detailed chronology of the human presence in the region. Directions for future research are clear. Parallel to an examination of the physical and biological characteristics of the palaeoenvironment, the close study of pseudo-tools will doubtless provide much discussion and food for thought. But clearly for each supposedly ancient site a multidisciplinary and broadly based study is absolutely necessary, and not only for the region discussed here.

-----§-----

Acknowledgements: The authors thank C. GUTH, O. BOEUF, J. DESSE and P.J. TEXIER who have provided access to the excavations at Chilhac; E. and M.F. BONIFAY for permitting a number of visits to the sites at Soleilhac, Nolhac and Ceyssaguet; F. CARRE for access to the material from Blassac and for his hospitality and assistance during the re-examination of Blassac - Les Battants in 1989; J. and D. CHOMETTE, D. LEFEVRE, T. PILLEYRE and G. VERNET for their collaboration in the field; J. COUTHURES for his unpublished information; G. CREVOLA for verifying the petrographic determinations and W. ROEBROEKS for reviewing the manuscript. European Science Fondation Meeting , Tautavel , sous presse - 12 -

References

Boèda E. 1990 - Qui étaient les premiers français ? La Recherche, n° 223, vol. 21, 96-968.

Boeuf O., 1983 - Le site villafranchien de Chilhac, en Haute-Loire. Etude paléontologique et biochronologique. Thèse de Doctorat Es-Sciences, Université Paris VII, 253 p.

Bonifay E. 1981 - Les traces des premiers Hominidés en France. La Recherche, 128, 12, 1442-1444.

Bonifay E. 1983 - La Préhistoire du Velay (Massif Central français) et le problème des origines de l'Homme. in La Recherche Archéologique en Haute-Loire, Annales du CRDP de Clermont-Ferrand, 9-16.

Bonifay E. 1989a- Un site du Très Ancien Paléolithique de plus de 2 M.a. dans le Massif Central français : Saint-Eble - Le Coupet (Haute-Loire). Comptes Rendus de l'Académie des Sciences de Paris, t. 308, série II, 1567-1570.

Bonifay E. 1989b - Les premiers signes de présence humaine. in Archéologie de la France : 30 ans de découvertes. Editions de la Réunion des Musées Nationaux, Paris, 36-37.

Bonifay E. 1989c- Paléolithique inférieur et moyen : premiers témoignages humains. in Archéologie de la France : 30 ans de découvertes. Ministère de la Culture, Editions de la Réunion des Musées Nationaux, Paris, 32-34.

Bonifay E. 1991 - Les premières industries du Sud-est de la France et du Massif Central. in Les premiers européens, CTHS Ed., Paris, 63-80.

Bonifay M.F. 1984 - Taphonomie quaternaire : Comparaison entre sites naturels et préhistoriques du Pléistocène moyen : gisements de Soleilhac (Haute-Loire). Journées d'étude de l'Association paléontologique française "Biocénoses et taphocénoses" , Paris, résumé, 18.

Bonifay M.F. 1987 - Biostratigraphie des grands mammifères d'Europe de l'Ouest.12° Congrès international de l'INQUA , Ottawa, Résumés : 133.

Bonifay E. et Bonifay M.F. 1981 - Le gisement préhistorique de Soleilhac (Blanzac, Haute-Loire). in Le bassin du Puy aux temps préhistoriques. Recherches récentes. le Puy, Musée Crozatier, 19-36.

Bonifay E. et Bonifay M.F. 1983 - Le Paléolithique ancien en Velay et en Auvergne : civilisations préhistoriques et milieu naturel. in Les Inédits de la Préhistoire Auvergnate, Musée Bargoin Clermont-Ferrand, p. 90-104.

Bonifay E., Mergoil J. 1988 - Les maars de Soleilhac (Blanzac, Haute-Loire) et leurs séries volcano-sédimentaires. Comptes Rendus de l'Académie des Sciences de Paris , t. 307, série II, 1561-1566.

Bonifay E., Consigny A., Liabeuf R. 1989 - Contribution du Massif Central français à la connaissance des premiers peuplements préhistoriques de l'Europe. Comptes Rendus de l'Académie des Sciences de Paris , t. 308, série II : 1491- 1496.

Boule M. 1889 - Temps quaternaires et préhistoriques du Cantal. Revue d'Anthropologie, p. 216.

Boule M. 1905 - L'origine des Eolithes. L'Anthropologie, 16, 257-267.

Boule M. 1921 - Les Hommes fossiles, Masson Ed.

Bourdier F. 1967 - Préhistoire de France. Nouvelle bibliothèque scientifique Flammarion, 412 p.

Bosinski G., Kröger K., Schäfer J., Turner E. 986 - Altsteinzeitliche Siedlungsplätze auf den Osteifel-Vulkanen. Jahrbuch des Römisch-Germanischen ZentralMuseums, 33, 97-130.

Bracco J.P. 1991 - Typologie, technologie et matières premières des industries du Très Ancien Paléolithique en Velay (Massif Central, France). Premiers résultats. in Les premiers européens, CTHS Ed., Paris, 93-100.

Breuil H. 1955 - Niveaux estuariens de galets fracturés par les vagues et les plages sableuses dans la basse Somme. Quaternaria, t. II, Rome, 21-29.

Bulle T. 1987 - Stat+, un utilitaire statistique. Archéologues et ordinateurs, 10, CNRS, 9-12.

Cantagrel J.M. et Baubron J.C. 1983 - Chronologie des éruptions dans le massif volcanique des Monts Dore : implications volcanologiques. Geol. Fr. (2) I, 1 (2), 123-142. European Science Fondation Meeting , Tautavel , sous presse - 13 -

Capitan L. et Marty P. 1924 - Les alluvions pontiennes des environs d'Aurillac (Cantal) et leurs silex taillés. Revue anthropologique, n° 3-4, 34è année, mars-avril 1924, 1-21.

Carré F. 1978 - Eléments d'industrie lithique dans les formations sédimentaires villafranchiennes de Blassac (Haute- Loire). Bulletin de la Société préhistorique française , t. 75, n° 5 : 131-132.

Carré F. 1983 - Le site de Blassac - Les Battants (haute-Loire). Fouilles 1971-1978. Eléments d'une stratigraphie. Congrés Préhistorique de France , 21è session, Cahors-Montauban, (1979), Société préhistorique française Ed., t.2., p 57-75.

Carré F. et Couthures J. 1982 - Nouveaux éléments de datation des formations volcaniques de Blassac (Haute-Loire, France) ; influences sur la chronologie des faunes "villafranchiennes" du Massif central français. 9è Réunion annuelle des Sciences de la Terre, Société géologique de France Ed., Paris : 121.

Carré F. 1991 - Les sites du Pliocène supérieur de Blassac - Haute-Loire (Massif Central, France). in Les premiers euro- péens, CTHS Ed., Paris, 135-142.

Chavaillon J. 1991 - Les ensembles lithiques de Chilhac III (Haute-Loire) : typologie, situation stratigraphique et analyse critique et comparative. in Les premiers européens, CTHS Ed., Paris, 81-91.

Clark J.D. 1958 - The natural fracture of pebbles from the Batoka Gorge, Northern Rhodesia, and its bearing on the kafuan industries of Africa. Proceedings of The Prehistoric Society, 24, 64-77.

Collina-Girard J. 1975 - Les industries archaïques sur galets des terrasses quaternaires de la plaine du Roussillon (Pyrénnées Orientales, France) . Thèse de 3° cycle, Université de Provence.

Collina-Girard J. 1986 - Grille descriptive et évolution typologique des industries archaïques : le modèle catalan. Bulletin de la Société préhistorique française , t. 83, n° 11-12 : 383-403.

Couthures J. 1982 - Contribution à la chronostratigraphie de formations plio-pléistocènes du Massif Central (France) . Thèse de 3° cycle, Paris VI, 2 vol.

Couthures J. et Pastre F. 1983 - Chronostratigraphie du Plio-pléistocène d'Auvergne et du Velay : nouveaux apports des datations radiométriques et du paléomagnétisme. Bulletin de l'Association française pour l'étude du Quaternaire , 1, 9-18.

Delson E. 1989 - Oldest Eurasian stone tools. Nature, 340, 96.

Diaz R. 1993 - Les premiers européens venaient-ils d'Afrique ? L'Histoire, 165, 42-45.

Fouris M. 1989 - Sites villafranchiens du Devès et les basaltes de la vallée de l'Allier : application de la méthode Potassium-Argon . Thèse d'Université, Clermont II.

Fournier A. 1971 - Méthode d'étude des industries archaïques sur galet. Diplôme d'Etudes Approfondies, Université de provence, 35 p.

Grayson D.K. 1986 - Eoliths, Archaeological Ambiguity, and the Generation of "Middle Range" Research. in D.J. Meltzer, D.D. Fowler, J.A. Sabloff Eds, American Archaeology Past and Future, Washington, Smithsonian Institution Press, 77-133.

G.U.E.R.P.A. 1984 - Présence possible d'Hominidés en Auvergne au Pliocène supérieur (2,5 M.a.) : l'apport des Etouaires (Issoire, Puy de Dôme). C. R. Acad. Sci., Paris, 299, 1091-1096.

Guth C. 1974 - Découverte dans le Villafranchien d'Auvergne de galets aménagés. C. R. Acad. Sci., Paris, 279, 1071-1072.

Guth C. et Boeuf O. 1977 - Les premiers galets aménagés du Villafranchien d'Auvergne. Le Courrier du CNRS, 25, 46-47.

Guth C. et Chavaillon J. 1985 - Découverte en 1984, de nouveaux outils paléolithiques à Chilhac III (Haute-Loire). Bulletin de la Société Préhistorique Française, 82, 56-64.

Haward F.W. 1911 - The Chipping of Flint by natural Agencies. Proceedings of the Prehistoric Society of East Anglia, 1, 185-193.

Haynes C.V. Jr. 1973 - The Calico Site : Artifacts or Geofacts ? Science, 181, 305-310.

Isaac G., Harris J.W.K., Marshall F. 1981 - Small is informative : the application of the study of mini-sites and least-effort criteria in the interpretation of the Early Pleistocene archaeological record at Koobi Fora, Kenya. in Las Industrias mas European Science Fondation Meeting , Tautavel , sous presse - 14 -

Antiguas, J.D. Clark and G. Isaac, X Congresso Union International de Ciencias Prehistoricas y Protohistoricas, Mexico (1981), 101-119.

Kulemeyer J. 1986 - Die alt-und mittelpaläolithischen Funde von Kärlich. Dissertation, Köln.

Le Gall O. et Raynal J.P. 1986 - Cerzat (Haute-Loire) : Le Pié du Roy, un site préhistorique dans un volcan. Revue Archéologique du Centre de la France, 25, 1, 99-100.

Lo Bello P., Féraud G., Hall C.M., York D., Lavina P., Bernat M. 1987 - 40Ar/39Ar step heating and laser fusion dating of a quaternary pumice from Neschers, Massif Central, France : the defeat of xenocrystic contamination. Chem. Geol. (Isot. geosci. Sect.), 66, 61-71.

Moir J.R. 1911 - The Natural Fracture of Flint and its Bearing Upon Rudimentary Flint Implements. Proceedings of the Prehistoric Society of East Anglia, 1, 171-18.

Mortelmans G. 1947 - Une cause d'erreur en préhistoire : la taille glaciaire. Bulletin de la Société Royale Belge d'Anthropologie et Préhistoire, t. LVIII, 60-71.

Peacok E. 1991 - Distinguishing between Artifacts and Geofacts: A Test Case from Eastern England. Journal of Field Archaeology, Vol. 18, 345-361.

Pei W.C. 1937 - Le rôle des phénomènes naturels dans l'éclatement et le façonnement des roches dures utilisées par l'homme préhistorique. Revue de Géographie Physique et de géologie Dynamique, 9 Paris, 349-423.

Pilleyre T. 1991 - Datation par thermoluminescence. Application à la chronologie des retombées volcaniques. Thèse de l'Université de Clermont II, DU345, 164 p.

Raynal J.P. 1987 - Evolution comparée de lacs de maars en Auvergne et Velay (France). Datation et contribution à la connaissance des climats pléistocènes. in Travaux français en Paléolimnologie, Documents du CERLAT, Mémoire n°1, 65- 96.

Raynal J.P. et Texier J.P. 1989 - Découverte d'Acheuléen ancien dans la carrière Thomas 1 à Casablanca et problème de l'ancienneté de la présence humaine au Maroc. Comptes Rendus de l'Académie des Sciences de Paris , t. 308, série II, 1743-1749.

Raynal J.P., Texier J.P., Geraads D., Alaoui F.Z. 1990 - Un nouveau gisement paléontologique plio-pléistocène en Afrique du Nord : Ahl Al Oughlam (ancienne carrière Déprez) à Casablanca (Maroc). Comptes Rendus de l'Académie des Sciences de Paris , t. 310, série II, 315-320.

Rio Carra M. 1991 - L'industrie sur quartz de Nolhac-Biard, un gisement du Très Ancien Paléolithique (résultats prélimi- naires). DEA, Université d'Aix-Marseille, 71 p., 14 pl.

Schnurrenberger D. and Alan L.B. 1985 - A Contribution to the Study of the Naturefact/Artifact Controversy. in M.G. Plew, J.C. Woods, M. G. Pavesic Eds, Stone Tool Analysis, Albuquerque, University of New Mexico Press, 133-159.

Tardy L. 1869 - Société d'Anthropologie de Paris, 16 décembre 1869.

Teulade A. 1985 - Etude téphrologique de formations lacustres de maars du Velay occidental. DEA, Université d'Aix- Marseille II, 72 p.

Teulade A. 1988 - Les tephra cendro-ponceux dans les sédiments du lac de maar de Soleilhac (Haute-Loire) : pétrographie et origine. Soc. Hist. Nat. Auvergne, 53, 39-43.

Teulade A. 1989 - Téphrologie des formations cendro-ponceuses en milieux lacustres quaternaires. Méthode d'étude et application au Massif Central français (Velay) et aux Carpathes orientales roumaines (dépression de Brassov). Thèse de l'Université d'Aix-Marseille II, 301 p.

Texier P.J. 1985 - Chilhac III : un gisement paléontologique villafranchien soliflué ? Bulletin de la Société préhistorique française, 82, 68-70.

Thouveny N. 1983 - Etude paléomagnétique de formations du Plio-pléistocène et de l'Holocène du Massif Central et de ses abords. Thèse, Université d'Aix-Marseille II, 132 p.

Thouveny N. et Bonifay E. 1984 - New chronological data on European Plio-Pleistocene faunas and hominid occupation sites. Nature, 308, n° 5957, 355-358. European Science Fondation Meeting , Tautavel , sous presse - 15 -

Villa P. 1991 - Middle Pleistocene Prehistory in Southwestern Europe: The State of Our Knowledge and Ignorance. Journal of Anthropological Research, 47, 2, 193-218.

Warren S.H. (1914) - The Experimental Investigation of Flint Fracture and its Application to Problems of Human Implements. Journal of the Royal Anthropological Institute of Great britain and Ireland, 44, 412-450.

Watson W. 1968 - Flint Implements, 3rd Ed. London. The Trustees of the British Museum. European Science Fondation Meeting , Tautavel , sous presse - 16 -

Figure 1 - Study area. European Science Fondation Meeting , Tautavel , sous presse - 17 -

TEPHROFACTS (122)

30 Chunks Pseudo-flakes 25 Pseudo-tools

20 % 15

0 Quartz Basalts Granites/Gneiss Flint Miscellaneous Petrography

Figure 2 - Petrography of tephrofacts. European Science Fondation Meeting , Tautavel , sous presse - 18 -

Figure 3 - EXAMPLES OF TEPHROFACTS (Arrows indicate clear directions of flaking) - Blassac-les Blanches Haute-Loire, pyroclastites - 1 : Partially cortical backed pseudo-flake and non cortical striking platform, fine grained granite. 2 : Partially cortical backed pseudo-flake and cortical striking platform, quartz. 3 : Partially cortical backed pseudo-flake and non cortical striking platform, fine grained granite. 4 : non-cortical backed pseudo-flake and non cortical striking platform, fine grained granite. 5 : Partially cortical backed pseudo-flake and cortical striking platform, pegmatitic quartz. 6 : pseudo-flake with non cortical edges and cortical striking platform, fine grained granite. 7 : pseudo-flake with non cortical edges and non cortical striking platform, pegmatitic quartz. 8 : Partially cortical backed pseudo-flake and cortical striking platform, fine grained granite. 9 : pseudo-flake with non cortical edges and non cortical striking platform, quartz. European Science Fondation Meeting , Tautavel , sous presse - 19 -

TYPES OF PSEUDO-FLAKES AND FLAKES

Tephrofacts (49) 50 Experimentation (97) Bourbonnais assemblages (372) 40 % 30

0 CSP-CB/PCB CSP-NCE NCSP-CB/PCB NCSP-NCE CSP: cortical striking platform. CB: cortical back. PCB: partially cortical back. NCE: non cortical edges. NCSP: non cortical striking platform.

Figure 4 - Types of pseudo-flakes compared to types of flakes from archaeological and experimental series.

WEIGHT OF PSEUDO-FLAKES AND FLAKES

100

60 Tephrofacts (49) Experimentation (100) % 50 Blassac - Les Battants (46) Nolhac (49) 40 Bourbonnais assemblages (371)

0 < 50g 51-100g 101-150g 151-200g 201-250g 251-300g > 301g Classes

Figure 5 - Weight of pseudo-flakes compared to those of archaeological and experimental series. European Science Fondation Meeting , Tautavel , sous presse - 20 -

Figure 6 - EXAMPLES OF TEPHROFACTS (Arrows indicate clear directions of flaking) - Miscellaneous from Haute-Loire - 1 : pseudo-chopper, basalte, Blassac-les Blanches pyroclastites. 2 : pseudo-chopper, lamprophyre, Blassac. 3 : pseudo- biface-like piece, basalte, Blassac-les Blanches pyroclastites. 4 : pseudo-chopper, basalte, Blassac-les!Blanches tuff-ring. 5 : pseudo-chopping-tool, migmatitic gneiss, Senèze tuff-ring inférieur. European Science Fondation Meeting , Tautavel , sous presse - 21 -

Figure 7 - EXAMPLES OF TEPHROFACTS (Arrows indicate clear directions of flaking) - Miscellaneous from Haute-Loire - 1 : peripherally flaked pseudo-chopper, granite, La Sauvetat tuff-ring. 2 : pseudo-polyhedron, pegmatitic quartz, Blassac- les Blanches pyroclastites. 3 : pseudo-invasively flaked tabular piece, granite, La Sauvetat tuff-ring. 4 : pseudo- polyhedron, granite, La Sauvetat tuff-ring. European Science Fondation Meeting , Tautavel , sous presse - 22 -

Figure 8 - EXAMPLES OF TEPHROFACTS (Arrows indicate clear directions of flaking) - Blassac-les Blanches Haute-Loire, pyroclastites - 1 : pseudo-partially bifacial discoïd piece, migmatitic gneiss à sillimanite. 2 : regular pseudo-polyhedron, quartz. 3 : regular pseudo-polyhedron, quartz. 4 : regular pseudo-polyhedron, oriented granite. 5 : pseudo-chopper, fine grained granite. European Science Fondation Meeting , Tautavel , sous presse - 23 -

Figure 9 - EXAMPLES OF TEPHROFACTS (Arrows indicate clear directions of flaking) - Miscellaneous from Haute-Loire - 1 : pseudo-invasively flaked tabular piece, granite, Blassac-les Blanches pyroclastites. 2 : pseudo-invasively flaked tabular piece, basalte, Blassac-les Blanches tuff-ring. 3 : pseudo-bifacially flaked discoïdal piece, basalte, Blassac-les Blanches tuff- ring. European Science Fondation Meeting , Tautavel , sous presse - 24 -

Figure 10 - EXAMPLES OF TEPHROFACTS (Arrows indicate clear directions of flaking) - Miscellaneous from Haute-Loire - 1 : pseudo-biface-like piece, granite d'anatexie, La Sauvetat tuff-ring. 2 : pseudo-chopper, basalte, La Sauvetat tuff-ring. 3 : pseudo-invasively sub-discoïd flaked tabular piece, oriented granite, Blassac-les Blanches pyroclastites. 4 : regular pseudo-polyhedron, fine grained granite, La Sauvetat tuff-ring. European Science Fondation Meeting , Tautavel , sous presse - 25 -

Figure 11 - EXAMPLES OF TEPHROFACTS (Arrows indicate clear directions of flaking) - Miscellaneous from Haute-Loire - 1 : pseudo-polyhedron, granite, La Sauvetat tuff-ring. 2 : pseudo-polyhedron, granite, Blassac-les Blanches pyroclastites. 3 : pseudo-polyhedron, granite, Blassac-les Blanches tuff-ring. 4 : pseudo-polyhedron, granite, La Sauvetat tuff-ring. European Science Fondation Meeting , Tautavel , sous presse - 26 -

COMPARISON WITH ASSEMBLAGES FROM TWO REPUTEDLY ANCIENT SITES

Tephrofacts (122) 70 Blassac - Les Battants (278) Chilhac III (45) 60

50 % 40

0 Quartz Basalt Granite/Gneiss Flint Miscellaneous Petrography

Figure 12 - Comparison with lithic series from two reputedly ancient sites.

COMPARISON WITH THREE ARCHAEOLOGICAL ASSEMBLAGES

100

Tephrofacts (122) 80 Nolhac (77) Soleilhac-Centre C (446) 70 Sainte-Anne 1 - J1 (2123)

% 50

0 Quartz Basalts/Phonolites Granites/Gneiss Flint Miscellaneous Petrography

Figure 13 - Comparison with three archaeological series.